Related Experiment Video
Updated: Sep 18, 2025
05:15
Detection of Neutralization-sensitive Epitopes in Antigens Displayed on Virus-Like Particle VLP-Based Vaccines Using a Capture Assay
Published on: February 10, 2022
3.7K
Utilizing Machine Learning to Improve Neutralization Potency of an HIV-1 Antibody Targeting the gp41 N-Heptad Repeat
Maria V Filsinger Interrante1,2,3, Shaogeng Tang3,4, Soohyun Kim3,4
1Stanford Biophysics Program, Stanford University School of Medicine, Stanford, California 94305, United States.
ACS Chemical Biology
|June 20, 2025
Summary
Researchers developed a potent HIV-1 N-heptad repeat (NHR) targeting antibody, D5_FI, using AI. This antibody shows broad neutralization against diverse HIV strains, offering new hope for HIV vaccines and therapies.
Area of Science:
- Immunology
- Virology
- Structural Biology
Background:
- The HIV-1 N-heptad repeat (NHR) in the gp41 prehairpin intermediate (PHI) is a conserved vaccine target.
- Potent NHR-targeting antibodies are lacking, hindering vaccine development and passive immunization strategies.
- Previous work yielded D5_AR, an improved NHR-directed monoclonal antibody (mAb) with tier-2 virus neutralization capabilities.
Purpose of the Study:
- To structurally characterize the D5_AR antibody bound to an NHR mimetic peptide.
- To engineer enhanced NHR-directed mAbs using protein language models and machine learning.
- To identify novel mAbs with improved neutralization potency against diverse HIV-1 strains.
Main Methods:
- Crystal structure determination of D5_AR complexed with NHR mimetic peptide IQN17 at 2.7Å resolution.
- Generation of small antibody variant libraries (<100 members) using protein language models and supervised machine learning.
- Screening of antibody variants for enhanced neutralization potency against HIV-1 pseudoviruses and replication-competent strains.
Main Results:
- The crystal structure of D5_AR bound to IQN17 was determined.
- A D5_AR variant, D5_FI, was identified with a 5-fold increase in neutralization potency.
- D5_FI demonstrates broad neutralization of tier-2 and tier-3 pseudoviruses, and R5/X4 replicating strains.
- Protein language models efficiently identified improved mAb variants from small libraries.
Conclusions:
- D5_FI represents the most potent NHR-directed monoclonal antibody characterized to date.
- AI-driven antibody engineering can rapidly generate highly effective therapeutic antibodies.
- This approach advances the development of NHR-based HIV-1 vaccines and antibody therapies.
